Dynamic demand systems have been employed in a number of studies to account for habit formation and inventory adjustments in demand. Few studies have attempted to provide a theoretical foundation for the dynamic demand structures employed. Recently, Bushehri (2003) showed how a generalized dynamic Rotterdam model could be derived from the neoclassical intertemporal utility maximization problem; however, no empirical application is provided in his study. This paper provides an empirical application of the generalized dynamic Rotterdam model to the demand for processed catfish products in the U.S. The two-period dynamic Rotterdam model explained a significant amount of the variation in U.S. catfish demand and was preferred to the one-period and static models. Estimates suggest that buyers adjust short-run inventories such that the past sales negatively affect current sales. Given inventory adjustment behavior, demand was relatively more inelastic in the long-run.